An alternative to viral mediated gene delivery is direct delivery of nucleic acid. This approach has several limitations including low efficiency of transfer, low stability and lack of cell specificity. In order to overcome some of these limitations other approaches have been made. These include non-specific ionic complexes with polycations such as polylysine (Wu and Wu, U.S. Pat. No. 5,166,320, contents of which are incorporated herein by reference) and histone. These bind non-specifically with the nucleic acid construct through polycations or basic proteins, such as histones. However, the resulting complexes still suffer some limitations including lack of uniformity of the complexes, lack of specificity with respect to polycation binding to specific regions of the nucleic acid construct, potential interference of complexes with nucleic acid and possible untimely dissassociation of the complex or lack of timely disassociation of the complex leading to a lack of stability of these nucleic acid polycation or nucleic polypeptide complexes.
Nucleic acid transfer to cells can take place by various methods. Such methods can utilize free nucleic acid, nucleic acid constructs or nucleic acid as part of the genome of a virus or bacteriophage vector.
Wu et al., U.S. Pat. No. 5,166,320, utilized a polynucleotide in a nonspecific association with the polycation polylysine. These complexes suffer limitations including lack of consistency of composition, lack of specificity with respect to polycation binding to specific regions of the nucleic acid construct, potential interference of complexes with nucleic acid and possible untimely dissassociation of the complex or lack of timely disassociation of the complex leading to a lack of stability of these nucleic acid or histone polycation or nucleic polypeptide complexes. This procedure does not provide for delivery of virus vectors. Furthermore, cell transformation efficiencies are still low.
Methods for retrovirus mediated gene transfer to hematopoietic cells ex vivo has been attempted in the presence of fibronectin or fibronectin fragments. Fibronectin binds to retroviruses but not to any other viruses, nucleic acids or nucleic acid constructs. Williams and Patel, WO 95/26200 (the contents of which are incorporated herein by reference), have transformed hematopoietic cells with retroviruses in the presence of fibronectin. The use of fibronectin in this way is limited only to use with some retrovirus vectors and not with other virus vectors or with nucleic acids.
It is desirable to form multimeric complexes for two primary reasons. The formation of such complexes results in an additive effect such that one can obtain collective activity of the monomeric units within a complex or these complexes could provide enhanced binding properties compared to the individual compounds or monomeric units, either through cooperative binding effects or through neighboring effects which produce higher localized concentrations. Polyligands usually exhibit higher binding affinities in the polymeric form than in the monomeric form as seen by the binding of polynucleotide sequences to their complementary sequences when compared to the binding of the monomeric units.
Multimeric complexes have been formed either by crosslinking of monomeric compounds directly or through a matrix or through the formation of noncovalent linkages. Examples of multimeric complexes formed by the crosslinking of a given compound such as enzymes, either directly or through a matrix are described in U.S. Pat. No. 4,687,732 (contents of which are incorporated herein by reference), whereby a visualization polymer composed of multiple units of a visualization monomer is linked together covalently by coupling agents which bond to chemical groups of the monomer. Examples of multimeric complexes made through the formation of noncovalent linkages such as ligand-receptor systems are the PAP (peroxidase-anti-peroxidase) complexes and APAAP (alkaline phosphatase-anti-alkaline phosphatase) complexes in common use as immunological reagents and the streptavidin-biotinylated enzyme complexes used for detection of biotinylated entities.
In the case of complexes formed by crosslinking or noncovalent binding, there are limitations with respect to the spacing and the chemical milieu of the monomeric unit within the complex which may affect the function and activity of the monomeric unit and as the size of the complex grows, solubility may be affected.
Efforts to regulate expression of procaryotic genes by eucaryotic processes have been attempted by Schwartz et al. (1993 Gene 127: 233) (also incorporated herein by reference) who introduced an intron sequence from a eucaryotic gene into a procaryotic gene. However, when introduced into a cell capable of mRNA processing, the gene expressed an altered protein in which additional amino acids were present due to the presence of flanking exon sequences associated with the inserted intron. This limitation is inherent in this approach since this method of intron isolation requires the a priori presence of inherent restriction sites in the exon regions flanking the intron, and intron insertion requires the presence of appropriate restriction sites in the gene receiving the intron. Therefore, even after the excision of the intron from the RNA, the flanking exon sequences remain as part of the coding sequence of the mature RNA molecules. Furthermore, the number of sites for intron insertion on the receiving gene is severely limited by the availability of appropriate restriction sites.
The alteration of the gene product by this approach may have unpredictable effects on the function of the gene product and severely limits the applicability of this method to specific instances. In the example of Schwartz et al. the additional amino acids had no apparent effect on the activity of the protein synthesized in the capable cell, but this is not always a predictable quality since it depends upon the site where the additional amino acids are incorporated. For instance, a short sequence coding for a small peptide introduced into the amino end of T7 RNA polymerase by Dunn et al. (1988 Gene 68: 259) (also incorporated herein by reference) had no apparent effect on enzyme activity. However introduction of the same sequence into a site near the carboxy terminus resulted in nearly complete loss of enzyme activity. Thus, the incorporation of extra amino acids as a result of introducing an exon into a coding sequence by the method of Schwartz et al. could have a drastic mutagenic effect.
Systems derived from procaryotic elements can produce functional products in mammalian cells. T7 RNA polymerase, an enzyme derived from an E. coli bacteriophage, has been expressed both transiently and stably in mammalian systems (Fuerst et al., 1986 Proc. Nat. Acad. Sci. U.S.A. 83: 8122, the contents of which are herein incorporated by reference). When synthesized in a mammalian environment, it is capable of acting upon genes under the control of a T7 promoter to produce transcripts that can be translated to provide a functional gene product. Large amounts of RNA can be transcribed from the T7 promoter (comprising up to 30,000 RNA molecules per cell, Lieber et al. 1993, also incorporated by reference).
In eucaryotic systems success has only been achieved by the use of a binary system with the polymerase on one construct and the T7 promoter on a separate construct, In this way either sequential transfections (Lieber et al. 1989 Nucleic Acids Res 17, 8485) (also incorporated by reference) or co-transfections with separate plasmids (Lieber et al. 1993 Methods Enzym. 217, 47) (incorporated by reference) or transfection with a plasmid containing a T7 promoter followed by infection with a recombinant vaccinia virus coding for T7 RNA polymerase (Fuerst et al. 1986 Proc. Nat. Acad. Sci. U.S.A. 83, 8122) must be done. Since T7 RNA polymerase can be cloned only free of a T7 promoter sequence (Davenloo et al. 1984 Proc. Nat. Acad. Sci. U.S.A. 81: 2035) (incorporated herein by reference), it appears that attempts to clone both elements in a single construct fail due to an event where synthesis of the T7 RNA polymerase-initiated transcription from the downstream promoter continues around the plasmid to direct more synthesis of T7 RNA polymerase leading to a cytocidal autocatalytic cascade. A similar strategy of elimination of cognate promoters has been described for the cloning of the bacterophage T3 (Morris et al. 1986 Gene 41: 193) and SP6 (Kotani et al. 1987 Nucl. Acids Res. 15: 2653) (both publications incorporated herein by reference) RNA polymerases. However, compatibility of these elements has been achieved by the addition of two modifications to the construct, i.e., inhibition of the T7 RNA polymerase by the presence of T7 lysozyme and the use of a repressible T7 lac promoter (Dubendorff and Studier 1991 J. Mol. Biol. 219: 61, 1991, incorporated herein by reference). Both of these limitations are required in order to obtain a construct.
The introduction of genetic material into cells can be done by two methods. One method is the exogenous application of nucleic acids which act directly on cellular processes but which themselves are unable to replicate or produce any nucleic acid. The intracellular concentrations of these molecules that must be achieved in order to affect cellular processes is dependent on the exogenous supply. Another method for nucleic acid delivery is the introduction into cells of Primary Nucleic Acid Constructs which themselves do not act on cellular processes but which produce single stranded nucleic acid in the cell which acts on cellular processes. In this case the introduced Primary Nucleic Acid Construct can integrate into cellular nucleic acid or it can exist in an extrachromosomal state, and it can propagate copies of itself in either the integrated or the extrachromosomal state. The nucleic acid construct can produce, from promoter sequences in the Primary Nucleic Acid Construct, single stranded nucleic acids which affect cellular processes of gene expression and gene replication. Such nucleic acids include antisense nucleic acids, sense nucleic acids and transcripts that can be translated into protein. The intracellular concentrantions of such nucleic acids are limited to promoter-dependent synthesis.
The effectiveness of single stranded nucleic acids produced from primary nucleic acid constructs is dependent on their concentration, the stability and the duration of production in the cell. Current methods for achieving intracellular concentrations are limited by a dependence on promoter directed synthesis.
The effectiveness of antisense therapy depends in large part on three major factors: a) the rate of transcription of antisense RNA, b) the cellular location of the RNA and c) the stability of the RNA molecules. While previous studies have addressed each of these factors, all three have not been addressed in a single approach. The present invention utilizes AS sequences substituted for nucleotide sequences in the U1 and other hnRNAs to achieve high nuclear concentrations of stable antisense RNA sequences.
U1, U2 and other snRNAs are nuclear-localized RNA molecules complexed with protein molecules. (Dahlberg and Lund 1988 in Structure and Function of Major and Minor Small Nuclear Ribonucleoprotein Particles, M. Birnstiel, Ed., Springer Verlag, Heidelberg, p 38: Zieve and Sautereau 1990 Biochemistry and Molecular Biology 25; 1, all of which are incorporated herein by reference).
The various promoters for U1, U2 and other snRNA operons are very strong and produce large amounts of RNA. U1 and other snRNAs have signals for export to the cytoplasm where specific proteins are complexed before reimportation to the cytoplasm as snRNPs (FIG. 41). snRNAs are very stable molecules. They form very highly ordered stem and loop structures (FIG. 43) which, when complexed with specific proteins, form snRNP, or splicesomes.
Antisense and other nucleic acid molecules which affect gene expression by acting on and altering RNA transcripts can derive certain advantages by confinement to the nucleus. Higher concentrations can be maintained in the smaller volume of the nucleus, interactions with target RNA can occur prior to their being used for expression and there would be no competition with messenger binding ribosomes.
Addition of antisense sequence to U2 RNA (Izant and Sardelli 1988 in Current Communications in Molecular Biology, Cold Spring Harbor, p 141, incorporated herein by reference) as a means of delivering antisense sequences altered the properties of normal U2 transcripts. Hybrid U2 molecules formed by insertion of antisense sequences into a restriction site in the 5′ end of the U2 transcript region showed decreasing antisense effectiveness with increasing insert size. Inserts longer than 250 bases substantially reduced antisense effectiveness. Furthermore, hybrids did not accumulate in the nucleus as efficiently as their wild type counterparts with the fraction of hybrids in the nucleus decreasing as insert length increased.
Yu and Weiner (1988 in Current Communications in Molecular Biology, Cold Spring Harbor, p 141, contents incorporated by reference) substituted 9 base antisense sequences directed at target sequences surrounding splice sites in mRNA. The antisense substitutions were made at the 5′end of U1 RNA. None of the antisense substitutions affected the level of targeted species of mature cytoplasmic RNA.
Constructs have been designed to increase antisense effectiveness by the inclusion of more than one targeting element in a single transcriptional unit. Multivalent constructs prepared in this way can produce numerous target directed entities acting on multiple target sites in nucleic acids. (Chen et al. 1992 in Antisense Strategies, Annals of the New York Academy of Sciences 660; 271: Zhow et al. Gene 1994 149; 33, both publications incorporated herein by reference). Different approaches to inhibition can be incorporated into a multivalent transcript as shown by Lisziewicz et al. (1993 Proc Natl Acad Sci USA 90, 8000, also incorporated by reference) who combined multiple copies of the HIV TAR with an antisense sequence to HIV gag on the same transcript.
The use of multivalent targeting by the inclusion of more than one targeting element on the same transcript provides a method for counteracting the high mutation rate of viruses such as HIV due to the unlikely event of simultaneous mutation of multiple target sequences. However, the common means of accomplishing these designs is the inclusion of the product entities on a single transcript. This approach suffers from the following limitations:
a) The total number of RNA molecules available as effective entities is limited by the strength of the single promoter;
b) During stable transformation of a cell, the integration event can disrupt the nucleic acid template sequence responsible for expression of the antisense sequence;
c) The use of multivalent transcripts is not favorable when one product entity present on the transcript acts on targets present in one cellular locale and another product entity present on the same multivalent transcript acts on targets present in a different cellular locale. This was the approach reported by Lisziewicz et al. (1993) where multiple TAR sequences, which act to bind the HIV tat protein in the cytoplasm, were present on the same transcript with antisense sequences for the HIV gag RNA, which are most effective in the nucleus.
Although there have been major efforts to find effective antiviral treatments, at the present time the only success has been in a diminution of virus growth rather than elimination of the virus. Among the efforts that have been pursued are attempts to prevent initiation of the virus replication cycle by preventing the virus from entering the cell by immunization or by treatment with antibodies or with proteins that interfere with virus recognition of a cell by interacting with the virus or the virus receptor site on the cell. These include unsuccessful treatment with high levels of soluble CD4 (Husson et al., 1992, incorporated by reference). In addition, efforts have been made to combat HIV infection after virus entry into a cell using protease inhibitors for preventing processing of viral polypeptides into functional proteins and varied nucleoside analogues which can block replication of the virus by inhibiting the activity of the virally encoded reverse transcriptase and other functions necessary for virus propagation. Stages of the processes of viral infection and viral replication cycle have been examined for the possibility of pharmacological or immunological intervention of the disease process. However, as independent and effective therapeutic agents, both immunological and small molecule inhibitors have failed to stem the progression of AIDS, and major problems remain in terms of effectiveness and the rise of viruses resistant to small molecule therapeutic agents. Even the application of combinations of immunological and small molecule agents has not been successful.
The introduction of genetic information into cells either to replace a function or to introduce a new function has provided an effective means for the treatment of viral infection. Genetic therapy approaches have been used to impart cell resistance to viruses by mechanisms which act intracellularly on the viral replication process (see Yu et al., Gene Therapy 1, 13-16 [1994, incorporated by reference). A result of these studies is that, in vitro, the effectiveness of genetic therapies is sensitive to virus concentration. Experiments in vitro that showed substantial levels of resistance at low ratios of virus to cells, at higher ratios showed a “breakthrough” phenomenon characterized by a period of seeming effectiveness followed by a surge in the virus production (Sczakiel et al. 1992 J. Virol 66; 5576: Scakiel and Pawlita 1990 J. Virol. 65; 468, all of which are incorporated by reference). Thus in vitro, at lower virus:cell ratios some genetically treated cells demonstrate longer survival times that at higher virus:cell ratios.
Compartmentalization of function is critical to regulated processes in eucaryotic cells. For example, the major part of cellular DNA is organized into chromosomes located in the nucleus where transcription of genetic information takes place. The major part of RNA synthesized in the nucleus is transported to the cytoplasm where it is translated. Other subcellular compartments for localized function include the Golgi apparatus, endoplasmic reticulum, nucleolus, mitochondria, chloroplast and the cellular membrane. Thus, a variety of mechanisms exist either to retain macromolecules in specific cellular compartments or to transport macromolecules from one cellular compartment to another. For example, in the directed exit of mature mRNA out of the nucleus into the cytoplasm, the presence of a 5′ cap, removal of introns and addition of a poly A sequence are all believed to contribute to the signal that directs the relocation (reference).
Some RNAs, such as small nuclear RNAs (snRNAs) involved in splicesome assembly, are relocated by sequential transportation (Dahlberg and Lund, 1988, in Structure and Function of Major and Minor Small Ribonuclear Particles, M. Birnstiel, ed., Springer Verlag, Heidelberg, pg. 38, incorporated by reference). After transcription in the nucleus, the presence of the 5′ cap and the processed 3′ terminus generate a bipartite signal for transport of U1 RNA into the cytoplasm. At this point there is further processing of the RNA by excision of a few nucleotides and hypermethylation of the 5′ cap. The binding of splicesome proteins present in the cytoplasm to the Sm region of the U1 RNA in combination with the hypermethylation is believed to generate a signal for the reimportation of the RNA back into the nucleus.
In contrast to most mRNA, most proteins do not need to be transported from their site of synthesis in the cytoplasm. However, some proteins that function in transcription, replication or other nuclear maintenance functions need to be present in the nucleus to function properly. In this case a polypeptide signal sequence present in the protein directs the transport of the protein from the cytoplasm into the nucleus. Still other proteins are not functional in the cytoplasm or in the nucleus but are required to be present in the membrane of the cell thereby requiring the presence of leader and lipophilic sequences.
The directing of target molecules as an approach to genetic therapy has been studied by attempts at localization for the express purpose of putting an active agent such as antisense RNA in proximity to the target in a particular cellular locale For example, some workers have designed nucleic acid constructs to express anti-sense RNA that would be retained in the nucleus in order to block newly transcribed target RNA from functioning (Izant and Sardelli, 1988, Current Communications in Molecular Biology, D. Melton, ed., Cold Spring Harbor Laboratory; Cotten and Birnstiel, 1989, EMBO Journal 8, 3861, incorporated by reference). The opposite effect has also been achieved by designing the transcript to include a signal for enhancing transport into the cytoplasm in order to block the translation of RNA that may be present there (Liszeiwicz et al. 1993, incorporated by reference).